Open chester-leung opened 4 years ago
Currently, the implementation requires that the aggregator participate in the federated learning by training on some of its own data. We should remove the need for the aggregator to have data.
Currently, the implementation requires that the aggregator participate in the federated learning by training on some of its own data. We should remove the need for the aggregator to have data.